While Bangladesh is gently stepping toward the next national election, social media and the growing strength of “artificial intelligence” algorithms are increasingly becoming the most powerful political forces influencing leaders, parties, and even coalitions.
The line between authentic political messaging and manufactured deception has become dangerously blurred with the rise of AI. In a country where over 50 million people use social media, platforms like Facebook, YouTube, and TikTok have become the new public squares. Even a moderate use of AI can amplify voices, mobilize opinions, and distort realities in ways that were unthinkable even a decade ago.
We have already seen how technology can tilt the scales. During the 2018 and 2024 so-called national elections, under the previous autocratic regime, the government weaponized social media platforms to amplify partisan narratives. The results were as dramatic as they were alarming: Viral misinformation, unauthorized call records, doctored images, and polarizing commentary flooded social networks.
PhotoIn Bangladesh’s political arena, where party leaders, supporters, and ordinary citizens tend to be less media literate, one false claim often spreads faster than fact and hits harder than truth. A fabricated story about an illegal act or a manipulated photo of opposition leaders can circulate across millions of feeds within hours, leaving lasting impressions even after debunking.
During the 2024 student-led protests, social media emerged as a powerful tool of empowerment and resistance, allowing students to organize, coordinate, and refuse mainstream media propaganda in real time. But the very platforms that gave student leaders visibility also became breeding grounds for false narratives. Student leaders were accused of accepting bribes or becoming politically compromised, while in reality they were forced by the government to call off the movement. These incidents explain the dual-edged nature of digital platforms. On one hand, they can be tools of freedom, and on the other, instruments of manipulation.
Now, the stakes are even higher with the emergence of AI-generated content. Deepfakes and AI-made propaganda aren’t some far-off threat; they are already here. We have all seen how quickly an image or clip can travel online. Now imagine a fake video of a major political figure -- generated using AI and blasted out through sponsored Facebook posts. It would not take long before doubts spread, reputations take a hit, and swing voters start to lean the other way.
We have already seen the use of AI-generated campaign songs, tailored messaging, and synthetic videos targeting local audiences in multiple languages in our neighboring countries like India. Robocalls mimicking President Joe Biden’s voice circulated during primaries earlier this year in the United States. Bangladesh is not immune; similar tactics could easily be deployed here, perhaps on an even larger scale due to weaker regulatory safeguards.
This is why Bangladesh stands at a crossroads. The upcoming election is not only about casting votes for the right leaders but also about how ordinary citizens safeguard themselves against the misinformation and manipulation of AI-driven content on social media. In a fragile democracy, the weaponization of technology could determine outcomes more decisively than manifestos, rallies, or debates.
The interim government has a historic opportunity to act. The answer does not lie in curbing speech or cracking down on dissent, which would only repeat the mistakes of the past. Rather, given the possible risks and threats in such a politically sensitive time, it should immediately create an independent, non-partisan commission dedicated to monitoring and addressing digital misinformation in real time. Such a body, staffed with technological and legal experts, could collaborate with platforms like Meta and Google to flag harmful content, demand accountability, and promote verified information.
Singapore and the European Union, for instance, have already established digital safety boards and laws mandating transparency in online political advertising. Bangladesh can adapt similar models suited to its own context.
Government action on its own will never be enough. Civil society, along with our universities, has to step in and take media literacy seriously. Young people may grow up online, but that doesn’t mean they truly understand how digital content works. Finland, for example, has built media literacy into its school system to guard against foreign disinformation, and the outcome so far has been promising. Bangladesh’s universities, NGOs, and journalists could take similar initiatives by launching workshops, campaigns, and fact-checking efforts that empower citizens to spot manipulation before it does harm.
Equally, social media companies must take their responsibility. For years, platforms have argued that algorithms are neutral tools, but in a politically fragile environment like Bangladesh, neutrality through inaction amounts to negligence. Facebook was heavily criticized after the Rohingya crisis in Myanmar, when hate speech spread unchecked. That lesson must not be forgotten. Global tech giants make profits from Bangladeshi users; they cannot turn away when their platforms are misused to undermine democratic processes.
Bangladesh is not alone in this struggle. In Brazil’s 2022 election, WhatsApp was flooded with misleading messages that shaped public debate. In the Philippines, TikTok and YouTube were misused to carry populist stories that decided how people voted. Bangladesh’s situation feels even especially urgent in light of its current political moment.
The fall of an autocratic regime has left the country at a rare point of political awakening -- and that makes it especially vulnerable. People are speaking more freely. They are organizing, debating, and demanding change. Yet they are also vulnerable -- to the seduction of virality, the certainty of fake outrage, and the emotions that AI can now mimic with eerie precision.
The future of our democracy may well be written in code, not ink. If we are not vigilant -- if we fail to build the tools, literacy, and ethics required to navigate this digital moment -- we may wake up after the election in a country we no longer recognize.
Md Kawsar Uddin, Associate Professor, Department of English and Modern Languages, IUBAT-- International University of Business Agriculture and Technology.